Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Anisotropic satellite galaxy quenching modulated by black hole activity


The evolution of satellite galaxies is shaped by their constant interaction with the circumgalactic medium surrounding central galaxies, which in turn may be affected by gas and energy ejected from the central supermassive black hole1,2,3,4,5,6. The nature of such a coupling between black holes and galaxies is, however, much debated7,8,9 and observational evidence remains scarce10,11. Here we report an analysis of archival data on 124,163 satellite galaxies in the potential wells of 29,631 dark matter halos with masses between 1012 and 1014 solar masses. We find that quenched satellite galaxies are relatively less frequent along the minor axis of their central galaxies. This observation might appear counterintuitive given that black hole activity is expected to eject mass and energy preferentially in the direction of the minor axis of the host galaxy. We show, however, that the observed anisotropic signal results precisely from the ejective nature of black hole feedback in massive halos, as outflows powered by active galactic nuclei clear out the circumgalactic medium, reducing the ram pressure and thus preserving star formation in satellite galaxies. This interpretation is supported by the IllustrisTNG suite of cosmological numerical simulations, even though the model’s sub-grid implementation of black hole feedback is effectively isotropic12.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Orientation of satellite galaxies around central galaxies.
Fig. 2: Anisotropic distribution of quiescent satellite galaxies in SDSS.
Fig. 3: SDSS versus IllustrisTNG.
Fig. 4: Anisotropic CGM density in IllustrisTNG.

Data availability

All data used in this work are publicly available through the Sloan Digital Sky Survey and the Illustris and IllustrisTNG public data releases.


  1. 1.

    Cicone, C. et al. Massive molecular outflows and evidence for AGN feedback from CO observations. Astron. Astrophys. 562, A21 (2014).

    Google Scholar 

  2. 2.

    Woo, J.-H., Bae, H.-J., Son, D. & Karouzos, M. The prevalence of gas outflows in type 2 AGNs. Astrophys. J. 817, 108 (2016).

    ADS  Google Scholar 

  3. 3.

    Pillepich, A. et al. Simulating galaxy formation with the IllustrisTNG model. Mon. Not. R. Astron. Soc. 473, 4077–4106 (2018).

    ADS  CAS  Google Scholar 

  4. 4.

    Nelson, D. et al. First results from the TNG50 simulation: galactic outflows driven by supernovae and black hole feedback. Mon. Not. R. Astron. Soc. 490, 3234–3261 (2019).

    ADS  CAS  Google Scholar 

  5. 5.

    Oppenheimer, B. D. et al. EAGLE and Illustris-TNG predictions for resolved eROSITA X-ray observations of the circumgalactic medium around normal galaxies. Astrophys. J. 893, L24 (2020).

    ADS  CAS  Google Scholar 

  6. 6.

    Davies, J. J., Crain, R. A., Oppenheimer, B. D. & Schaye, J. The quenching and morphological evolution of central galaxies is facilitated by the feedback-driven expulsion of circumgalactic gas. Mon. Not. R. Astron. Soc. 491, 4462–4480 (2020).

    ADS  CAS  Google Scholar 

  7. 7.

    Harrison, C. M. et al. AGN outflows and feedback twenty years on. Nat. Astron. 2, 198–205 (2018).

    ADS  Google Scholar 

  8. 8.

    Dashyan, G. et al. AGN-driven quenching of satellite galaxies. Mon. Not. R. Astron. Soc. 487, 5889–5901 (2019).

    ADS  CAS  Google Scholar 

  9. 9.

    Veilleux, S., Maiolino, R., Bolatto, A. D. & Aalto, S. Cool outflows in galaxies and their implications. Astron. Astrophys. Rev. 28, 2 (2020).

    ADS  Google Scholar 

  10. 10.

    Cheung, E. et al. Suppressing star formation in quiescent galaxies with supermassive black hole winds. Nature 533, 504–508 (2016).

    ADS  CAS  PubMed  Google Scholar 

  11. 11.

    Martín-Navarro, I., Brodie, J. P., Romanowsky, A. J., Ruiz-Lara, T. & van de Ven, G. Black-hole-regulated star formation in massive galaxies. Nature 553, 307–309 (2018).

    ADS  PubMed  Google Scholar 

  12. 12.

    Weinberger, R. et al. Simulating galaxy formation with black hole driven thermal and kinetic feedback. Mon. Not. R. Astron. Soc. 465, 3291–3308 (2017).

    ADS  CAS  Google Scholar 

  13. 13.

    Tempel, E. et al. Flux- and volume-limited groups/clusters for the SDSS galaxies: catalogues and mass estimation. Astron. Astrophys. 566, A1 (2014).

    Google Scholar 

  14. 14.

    Ahn, C. P. et al. The Tenth Data Release of the Sloan Digital Sky Survey: first spectroscopic data from the SDSS-III Apache Point Observatory Galactic Evolution Experiment. Astrophys. J. 211 (Suppl.), 17 (2014).

    Google Scholar 

  15. 15.

    Kauffmann, G. et al. Stellar masses and star formation histories for 105 galaxies from the Sloan Digital Sky Survey. Mon. Not. R. Astron. Soc. 341, 33–53 (2003).

    ADS  Google Scholar 

  16. 16.

    Brinchmann, J. et al. The physical properties of star-forming galaxies in the low-redshift Universe. Mon. Not. R. Astron. Soc. 351, 1151–1179 (2004).

    ADS  CAS  Google Scholar 

  17. 17.

    Stoughton, C. et al. Sloan Digital Sky Survey: early data release. Astron. J. 123, 485–548 (2002).

    ADS  Google Scholar 

  18. 18.

    Nelson, D. et al. The IllustrisTNG simulations: public data release. Comput. Astrophys. Cosmol. 6, 2 (2019).

    ADS  Google Scholar 

  19. 19.

    Rodriguez-Gomez, V. et al. The optical morphologies of galaxies in the IllustrisTNG simulation: a comparison to Pan-STARRS observations. Mon. Not. R. Astron. Soc. 483, 4140–4159 (2019).

    ADS  CAS  Google Scholar 

  20. 20.

    Donnari, M. et al. Quenched fractions in the IllustrisTNG simulations: comparison with observations and other theoretical models. Preprint at (2020).

  21. 21.

    Fujita, Y. Pre-processing of galaxies before entering a cluster. Publ. Astron. Soc. Jpn 56, 29–43 (2004).

    ADS  Google Scholar 

  22. 22.

    Kauffmann, G., Li, C., Zhang, W. & Weinmann, S. A re-examination of galactic conformity and a comparison with semi-analytic models of galaxy formation. Mon. Not. R. Astron. Soc. 430, 1447–1456 (2013).

    ADS  Google Scholar 

  23. 23.

    Nelson, D. et al. The illustris simulation: public data release. Astron. Comput. 13, 12–37 (2015).

    ADS  Google Scholar 

  24. 24.

    Davé, R. et al. SIMBA: cosmological simulations with black hole growth and feedback. Mon. Not. R. Astron. Soc. 486, 2827–2849 (2019).

    ADS  Google Scholar 

  25. 25.

    McNamara, B. R. & Nulsen, P. E. J. Heating hot atmospheres with active galactic nuclei. Annu. Rev. Astron. Astrophys. 45, 117–175 (2007).

    ADS  Google Scholar 

  26. 26.

    Fabian, A. C. Observational evidence of active galactic nuclei feedback. Annu. Rev. Astron. Astrophys. 50, 455–489 (2012).

    ADS  CAS  Google Scholar 

  27. 27.

    Gunn, J. E., Gott, I. & Richard, J. On the infall of matter into clusters of galaxies and some effects on their evolution. Astrophys. J. 176, 1 (1972).

    ADS  Google Scholar 

  28. 28.

    Yun, K. et al. Jellyfish galaxies with the IllustrisTNG simulations—I. Gas-stripping phenomena in the full cosmological context. Mon. Not. R. Astron. Soc. 483, 1042–1066 (2019).

    ADS  CAS  Google Scholar 

  29. 29.

    Maiolino, R. et al. Star formation inside a galactic outflow. Nature 544, 202–206 (2017).

    ADS  CAS  PubMed  Google Scholar 

  30. 30.

    Navarro, J. F., Frenk, C. S. & White, S. D. M. The structure of cold dark matter halos. Astrophys. J. 462, 563 (1996).

    ADS  CAS  Google Scholar 

  31. 31.

    Martín-Navarro, I., Burchett, J. N. & Mezcua, M. Quantifying the effect of black hole feedback from the central galaxy on the satellite populations of groups and clusters. Astrophys. J. 884, L45 (2019).

    ADS  Google Scholar 

  32. 32.

    de Vaucouleurs, G. Recherches sur les nebuleuses extragalactiques. Ann. Astrophys. 11, 247 (1948).

    ADS  Google Scholar 

  33. 33.

    Baes, M. et al. Efficient three-dimensional NLTE dust radiative transfer with SKIRT. Astrophys. J. 196 (Suppl.), 22 (2011).

    Google Scholar 

  34. 34.

    Camps, P., Baes, M. & Saftly, W. Using 3D Voronoi grids in radiative transfer simulations. Astron. Astrophys. 560, A35 (2013).

    ADS  Google Scholar 

  35. 35.

    Donnari, M. et al. The star formation activity of IllustrisTNG galaxies: main sequence, UVJ diagram, quenched fractions, and systematics. Mon. Not. R. Astron. Soc. 485, 4817–4840 (2019).

    ADS  CAS  Google Scholar 

  36. 36.

    Nelson, D. et al. First results from the IllustrisTNG simulations: the galaxy colour bimodality. Mon. Not. R. Astron. Soc. 475, 624–647 (2018).

    ADS  CAS  Google Scholar 

  37. 37.

    Huertas-Company, M. et al. The Hubble Sequence at z ~ 0 in the IllustrisTNG simulation with deep learning. Mon. Not. R. Astron. Soc. 489, 1859–1879 (2019).

    ADS  CAS  Google Scholar 

  38. 38.

    Zanisi, L. et al. A deep learning approach to test the small-scale galaxy morphology and its relationship with star formation activity in hydrodynamical simulations. Mon. Not. R. Astron. Soc. 501, 4359–4382 (2020).

    ADS  Google Scholar 

  39. 39.

    Foreman-Mackey, D., Hogg, D. W., Lang, D. & Goodman, J. emcee: The MCMC hammer. Publ. Astron. Soc. Pacif. 125, 306 (2013).

    ADS  Google Scholar 

  40. 40.

    Simard, L., Mendel, J. T., Patton, D. R., Ellison, S. L. & McConnachie, A. W. A catalog of bulge+disk decompositions and updated photometry for 1.12 million galaxies in the Sloan Digital Sky Survey. Astrophys. J. 196 (Suppl.), 11 (2011).

    Google Scholar 

  41. 41.

    Martín-Navarro, I., Burchett, J. N. & Mezcua, M. Black hole feedback and the evolution of massive early-type galaxies. Mon. Not. R. Astron. Soc. 491, 1311–1319 (2020).

    ADS  Google Scholar 

  42. 42.

    van den Bosch, R. C. E. Unification of the fundamental plane and super massive black hole masses. Astrophys. J. 831, 134 (2016).

    ADS  Google Scholar 

  43. 43.

    Martín-Navarro, I., Brodie, J. P., van den Bosch, R. C. E., Romanowsky, A. J. & Forbes, D. A. Stellar populations across the black hole mass-velocity dispersion relation. Astrophys. J. 832, L11 (2016).

    ADS  Google Scholar 

  44. 44.

    Terrazas, B. A. et al. Quiescence correlates strongly with directly measured black hole mass in central galaxies. Astrophys. J. 830, L12 (2016).

    ADS  Google Scholar 

  45. 45.

    Terrazas, B. A., Bell, E. F., Woo, J. & Henriques, B. M. B. Supermassive black holes as the regulators of star formation in central galaxies. Astrophys. J. 844, 170 (2017).

    ADS  Google Scholar 

  46. 46.

    Terrazas, B. A. et al. The relationship between black hole mass and galaxy properties: Examining the black hole feedback model in IllustrisTNG. Preprint at (2019).

  47. 47.

    Dullo, B. T., Bouquin, A. Y. K., Gil De Paz, A., Knapen, J. H. & Gorgas, J. The (black hole mass)-(color) relations for early- and late-type galaxies: red and blue sequences. Preprint at (2020).

  48. 48.

    Li, Y. et al. Correlations between black holes and host galaxies in the Illustris and IllustrisTNG Simulations. Astrophys. J. 895, 102 (2020).

    ADS  CAS  Google Scholar 

  49. 49.

    Donnari, M. et al. Quenched fractions in the IllustrisTNG simulations: the roles of AGN feedback, environment, and pre-processing. Preprint at (2020).

  50. 50.

    Genel, S. et al. Introducing the Illustris project: the evolution of galaxy populations across cosmic time. Mon. Not. R. Astron. Soc. 445, 175–200 (2014).

    ADS  CAS  Google Scholar 

  51. 51.

    Kauffmann, G. et al. The morphology and kinematics of the gaseous circumgalactic medium of Milky Way mass galaxies—II. Comparison of IllustrisTNG and Illustris simulation results. Mon. Not. R. Astron. Soc. 486, 4686–4700 (2019).

    ADS  CAS  Google Scholar 

  52. 52.

    Lim, S. H. et al. Properties of the CGM and IGM: constraints on galaxy formation models from the Sunyaev-Zel’dovich effect. Preprint at (2020).

  53. 53.

    Pillepich, A. et al. First results from the IllustrisTNG simulations: the stellar mass content of groups and clusters of galaxies. Mon. Not. R. Astron. Soc. 475, 648–675 (2018).

    ADS  CAS  Google Scholar 

Download references


I.M.-N. acknowledges support from grant PID2019-107427GB-C32 from The Spanish Ministry of Science and Innovation and from the Marie Skłodowska-Curie Individual SPanD Fellowship 702607. A.P. and M.D. acknowledge support by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Project-ID 138713538 – SFB 881 (“The Milky Way System”), subproject A01. We thank G. Pérez Díaz for helping with the design of the figures.

Author information




I.M.-N. and A.P. developed the original idea and characterized the signal in the observed and simulated data. D.N. measured the gas mass density distribution in IllustrisTNG and contributed to the early developement of the project. V.R.-G. generated the synthetic SDSS-like images based on IllustrisTNG data, and M.D. provided the information about the infalling time of satellites in IllustrisTNG. L.H. and V.S. contributed to the analysis and interpretation of the observed and simulated data. I.M.-N. and A.P. wrote the text, and all the co-authors contributed to refining and polishing the final manuscript.

Corresponding author

Correspondence to Ignacio Martín-Navarro.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Claudia Cicone and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 SDSS posterior distributions for the best-fitting description of the angular modulation of satellite quiescence.

We fit the observed data with a cosine function with three free parameters, the average quiescent fraction fq, the amplitude of the modulation, and a re-scaling term for the expected error f. Posteriors are well behaved and allowed us to reject the null-hypothesis at a ~6σ level. Blue solid vertical lines indicate the best-fitting values and the dashed lines indicate the 1σ confidence interval.

Extended Data Fig. 2 Sensitivity of the SDSS signal to PA uncertainties.

a, The fraction of SDSS quiescent galaxies as a function of the orientation based on the worst-fitting functional form (de Vaucouleurs versus exponential) according to the SDSS photometric pipeline. The stability of the signal demonstrates that our results are robust against the photometric fitting procedure. Error bars represent the best-fitting standard deviation, as described in the Methods. b, Coloured curves indicate the best-fitting solution for SDSS data obtained while randomly perturbating the PA of the central galaxy by ΔPA. For reference, black symbols and curves are the same as in Fig. 2. A clear modulation in the fraction of quiescent galaxies is observed even for ΔPA ≈ 30, which is an order of magnitude larger than the expected error on the individual PAs. c, d, The SDSS g-band images of galaxies best-fitted by a de Vaucouleurs (top row) and an exponential profile (bottom row), with the PA uncertainty indicated by the white-shaded area. The adopted PA is indicated in the top left corner of each image.

Extended Data Fig. 3 Test with randomized PAs.

a, The fraction of quiescent satellites in SDSS data after randomizing the PA of the central galaxies. As expected, no signal is recovered in this case. Error bars represent the best-fitting standard deviation, as described in the Methods. b, The posterior distributions for this test, where the modelled amplitude is consistent with no angular variation. Blue solid vertical lines indicate the best-fitting values and the dashed lines indicate the 1σ confidence interval.

Extended Data Fig. 4 Characterization of the SDSS signal.

a, We show that the modulation in the observed signal is higher for satellites closer to the centre (Rsat < 0.5Rvir, orange symbols) than for those satellites in the outskirts (Rsat < 0.5Rvir, blue symbols). b, The signal is stronger for halos with more massive central galaxies (logMcen > 11M, orange symbols) compared to the signal observed in halos with less massive central galaxies (logMcen < 11M, blue symbols). c, Less massive satellites (logMsat < 10.5M, orange symbols) exhibit a larger variation than more massive ones (logMsat > 10.5M, blue symbols). d, The signal is also stronger in halos hosting more massive black holes in their centre (orange symbols), compared to those with relatively less-massive central black holes (blue symbols) Panels eh are equivalent to panels ad but without removing the offset between the different sub-samples.

Extended Data Fig. 5 Alternative metrics for the characterization of SDSS satellites’ star-formation status.

a, b, The modulation observed in the average specific SFR (a) and distance from the star-formation main sequence (b) of SDSS satellites closely follows that shown by the fraction of quiescent satellites in Fig. 2. Regardless of the metric used to characterize the star-formation properties of satellite galaxies, there is a clear dependence on the orientation with respect to the central galaxy. Error bars indicate the 1σ uncertainty and yellow lines mark the location of the minor and major axes.

Extended Data Fig. 6 Additional trends with halo mass and distance in SDSS.

As in Fig. 2, black symbols represent the observed modulation on the SDSS data. The blue line indicates the change in the quiescent fraction that could be expected because of the average halo mass dependence on orientation, which is much smaller than the reported one. Similarly, satellites along the minor axis are marginally closer to the central galaxy than along the major axis, leading to a negative and even weaker modulation, as shown by the red line. Error bars represent the best-fitting standard deviation, as described in the Methods.

Extended Data Fig. 7 Iso-quiescent fraction contours.

Similarly to Fig. 4, the contours of constant fq are shown, but this time at three different levels: fq = {0.36, 0.42, 0.48}. The background image corresponds to the IllustrisTNG gas over-density and the typical virial radius in the explored halo mass range is shown as a dashed grey circle, as in Fig. 4.

Extended Data Fig. 8 IllustrisTNG versus Illustris comparison.

Modulation in the fraction of quiescent galaxies for the IllustrisTNG (namely, TNG100, red symbols) and the original Illustris (blue symbols) simulations. Error bars represent the best-fitting standard deviation, as described in the Methods. The signal is shown in green for a sample of IllustrisTNG satellites with the same mass distribution as those in Illustris, to assess the possible effect of a mass bias between the two simulations. Both simulations probe a similar ~100-Mpc comoving cosmological volume and thus share the same large-scale structure properties; the treatment of black hole growth and feedback is the most relevant difference between the two. However, it is clear that the amplitude of the modulation is much higher in IllustrisTNG (0.032 ± 0.004) than in Illustris (0.013 ± 0.007).

Extended Data Fig. 9 Quiescent versus star-forming central galaxies in IllustrisTNG and SDSS.

In a, at a fixed central stellar mass of ~logMcen = 10.5M, the modulation in the fraction of quiescent satellites in TNG100 is shown for star-forming (blue) and quiescent (orange) central galaxies. Although there are a limited number of satellites, the modulation in the signal appears to be stronger for quiescent central galaxies than for star-forming ones. Since quiescentness in IllustrisTNG is a strong indication of an effective black hole feedback, the fact that the signal is stronger for quiescent galaxies is also an indication of the proposed AGN-related origin for the observed quenching directionality. The modulation in the fraction of quiescent satellites is shown for star-forming (blue) and quiescent (orange) central galaxies in b but this time for SDSS galaxies, again of logMcen = 10.5M. The observed modulation is stronger for quiescent than for star-forming central galaxies as seen in IllustrisTNG. Solid lines and shaded areas indicate the best-fitting trends and 1σ confidence interval, respectively. Error bars represent the best-fitting standard deviation, as described in the Methods.

Extended Data Fig. 10 Dependencies of the signal in IllustrisTNG.

a, The fraction of quiescent satellites around central galaxies whose black holes have injected, relatively to their mass, more (red) and less (blue) total energy. b, Similarly, the same separation but in this case considering only the kinetic energy injected by the black holes. In both cases, the amplitude of the modulation is stronger when the total (a) and kinetic (b) energy released by the central black holes increase. Similar to Extended Data Fig. 4, panel c shows how the signal in IllustrisTNG depends on the relative mass of the central black hole, being stronger for more over-massive black hole galaxies. d, The observed signal in IllustrisTNG (red) and the de-projected signal (blue) using the underlying 3D satellite distribution. We note that in d we did not impose any cut in central stellar mass and therefore absolute values are different from the other panels. Error bars and shaded areas represent 1σ confidence intervals, and solid lines are the best-fitting solutions.

Extended Data Fig. 11 Quenching directionality in IllustrisTNG.

a, The number of TNG100 satellites in each orientation bin, depending on whether they are star-forming (blue symbols), quenched in their z ≈ 0 host halo (green), were pre-processed and quenched in a different halo (orange), or quenched as central galaxies (red). The last two groups (red and orange symbols) are sensitive to large-scale structure effects, but correspond only to a small fraction of the total satellite population. b, The fraction of quiescent satellites as a function of orientation is shown but only for those satellites that quenched in their z ≈ 0 host halo (green symbols). The amplitude of this modulation mimics that measured for all IllustrisTNG satellites (grey-shaded area and black line).

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Martín-Navarro, I., Pillepich, A., Nelson, D. et al. Anisotropic satellite galaxy quenching modulated by black hole activity. Nature 594, 187–190 (2021).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing